In [1]:
data_dir = './data'
import helper
helper.download_extract('mnist', data_dir)
helper.download_extract('celeba', data_dir)
In [2]:
show_n_images = 25
%matplotlib inline
import os
from glob import glob
from matplotlib import pyplot
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'mnist/*.jpg'))[:show_n_images], 28, 28, 'L')
pyplot.imshow(helper.images_square_grid(mnist_images, 'L'), cmap='gray')
Out[2]:
CelebA¶
一个包含 20 多万张名人图片及相关图片说明的数据集
In [3]:
show_n_images = 25
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'img_align_celeba/*.jpg'))[:show_n_images], 28, 28, 'RGB')
pyplot.imshow(helper.images_square_grid(mnist_images, 'RGB'))
Out[3]:
In [4]:
from distutils.version import LooseVersion
import warnings
import tensorflow as tf
# Check TensorFlow Version
assert LooseVersion(tf.__version__) >= LooseVersion('1.0'), 'Please use TensorFlow version 1.0 or newer. You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))
# Check for a GPU
if not tf.test.gpu_device_name():
warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))
输入（Input）¶
In [5]:
def model_inputs(image_width, image_height, image_channels, z_dim):
"""
Create the model inputs
:param image_width: The input image width
:param image_height: The input image height
:param image_channels: The number of image channels
:param z_dim: The dimension of Z
:return: Tuple of (tensor of real input images, tensor of z data, learning rate)
"""
input_real = tf.placeholder(tf.float32,[None, image_width, image_height, image_channels])
input_z = tf.placeholder(tf.float32, [None, z_dim])
learning_rate = tf.placeholder(tf.float32)
return input_real, input_z, learning_rate
辨别器（Discriminator）¶
In [6]:
def discriminator(images, reuse=False):
"""
Create the discriminator network
:param image: Tensor of input image(s)
:param reuse: Boolean if the weights should be reused
:return: Tuple of (tensor output of the discriminator, tensor logits of the discriminator)
"""
with tf.variable_scope('discriminator', reuse=reuse):
alpha = 0.1
dropout = 0.6
#28*28*3
x1 = tf.layers.conv2d(images, 64, 5, strides=2 ,padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
relu1 = tf.maximum(alpha * x1, x1)
drop1 = tf.nn.dropout(relu1, dropout)
#14*14*64
x2 = tf.layers.conv2d(drop1, 128, 5, strides=2 ,padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
bn2 = tf.layers.batch_normalization(x2, training=True)
relu2 = tf.maximum(alpha * bn2, bn2)
drop2 = tf.nn.dropout(relu2, dropout)
#7*7*128
x3 = tf.layers.conv2d(drop2, 256, 5, strides=2 ,padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
bn3 = tf.layers.batch_normalization(x3, training=True)
relu3 = tf.maximum(alpha * bn3, bn3)
drop3 = tf.nn.dropout(relu3, dropout)
#4*4*256
# x4 = tf.layers.conv2d(drop3, 512, 5, strides=1, padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
# bn4 = tf.layers.batch_normalization(x4, training=True)
# relu4 = tf.maximum(alpha * bn4, bn4)
# drop4 = tf.nn.dropout(relu4, dropout)
#4*4*512
flat = tf.reshape(drop3, [-1, 4*4*256])
logits = tf.layers.dense(flat, 1, kernel_initializer=tf.contrib.layers.xavier_initializer())
out = tf.nn.sigmoid(logits)
return out, logits
生成器（Generator）¶
In [7]:
def generator(z, out_channel_dim, is_train=True):
"""
Create the generator network
:param z: Input z
:param out_channel_dim: The number of channels in the output image
:param is_train: Boolean if generator is being used for training
:return: The tensor output of the generator
"""
with tf.variable_scope('generator', reuse=not is_train):
alpha = 0.1
dropout = 0.6
#fully connected and reshape
x1 = tf.layers.dense(z, 7*7*256, kernel_initializer=tf.contrib.layers.xavier_initializer())
x1 = tf.layers.batch_normalization(x1, training=is_train)
x1 = tf.nn.relu(x1)
x1 = tf.reshape(x1, [-1, 7, 7, 256])
x1 = tf.nn.dropout(x1, dropout)
#7*7*256
x2 = tf.image.resize_nearest_neighbor(x1, (14, 14))
x2 = tf.layers.conv2d_transpose(x2, 128, 5, strides=1, padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
x2 = tf.layers.batch_normalization(x2, training=is_train)
x2 = tf.nn.relu(x2)
#n14*14*128
x3 = tf.image.resize_nearest_neighbor(x2, (28, 28))
x3 = tf.layers.conv2d_transpose(x3, 64, 5, strides=1, padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
x3 = tf.layers.batch_normalization(x3, training=is_train)
x3 = tf.nn.relu(x3)
#28*28*64
# x4 = tf.layers.conv2d_transpose(x3, 32, 5, strides=1, padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
# x4 = tf.layers.batch_normalization(x4, training=is_train)
# x4 = tf.nn.relu(x4)
#28*28*32
#output layer
logits = tf.layers.conv2d_transpose(x3, out_channel_dim, 5, strides=1, padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
out = tf.nn.tanh(logits)
return out
损失函数（Loss）¶
In [8]:
def model_loss(input_real, input_z, out_channel_dim):
"""
Get the loss for the discriminator and generator
:param input_real: Images from the real dataset
:param input_z: Z input
:param out_channel_dim: The number of channels in the output image
:return: A tuple of (discriminator loss, generator loss)
"""
g_model = generator(input_z, out_channel_dim)
d_model_real, d_logits_real = discriminator(input_real)
d_model_fake, d_logits_fake = discriminator(g_model, reuse=True)
d_loss_real = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real, labels=tf.ones_like(d_model_real) * 0.9))
d_loss_fake = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.zeros_like(d_model_fake)))
d_loss = d_loss_real + d_loss_fake
g_loss = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.ones_like(d_model_fake)))
return d_loss, g_loss
优化（Optimization）¶
In [9]:
def model_opt(d_loss, g_loss, learning_rate, beta1):
"""
Get optimization operations
:param d_loss: Discriminator loss Tensor
:param g_loss: Generator loss Tensor
:param learning_rate: Learning Rate Placeholder
:param beta1: The exponential decay rate for the 1st moment in the optimizer
:return: A tuple of (discriminator training operation, generator training operation)
"""
# TODO: Implement Function
t_vars = tf.trainable_variables()
g_vars = [var for var in t_vars if var.name.startswith('generator')]
d_vars = [var for var in t_vars if var.name.startswith('discriminator')]
with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)):
d_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(d_loss, var_list=d_vars)
g_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(g_loss, var_list=g_vars)
return d_train_opt, g_train_opt
In [10]:
import numpy as np
def show_generator_output(sess, n_images, input_z, out_channel_dim, image_mode):
"""
Show example output for the generator
:param sess: TensorFlow session
:param n_images: Number of Images to display
:param input_z: Input Z Tensor
:param out_channel_dim: The number of channels in the output image
:param image_mode: The mode to use for images ("RGB" or "L")
"""
cmap = None if image_mode == 'RGB' else 'gray'
z_dim = input_z.get_shape().as_list()[-1]
example_z = np.random.uniform(-1, 1, size=[n_images, z_dim])
samples = sess.run(
generator(input_z, out_channel_dim, False),
feed_dict={input_z: example_z})
images_grid = helper.images_square_grid(samples, image_mode)
pyplot.imshow(images_grid, cmap=cmap)
pyplot.show()
训练¶
In [11]:
def train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode):
"""
Train the GAN
:param epoch_count: Number of epochs
:param batch_size: Batch Size
:param z_dim: Z dimension
:param learning_rate: Learning Rate
:param beta1: The exponential decay rate for the 1st moment in the optimizer
:param get_batches: Function to get batches
:param data_shape: Shape of the data
:param data_image_mode: The image mode to use for images ("RGB" or "L")
"""
# Build Model
input_real, input_z, _ = model_inputs(data_shape[1], data_shape[2], data_shape[3], z_dim)
d_loss, g_loss = model_loss(input_real, input_z, data_shape[3])
global_step = tf.Variable(0, trainable=False)
d_train_opt, _ = model_opt(d_loss, g_loss, learning_rate, beta1)
_, g_train_opt = model_opt(d_loss, g_loss, learning_rate, beta1)
saver = tf.train.Saver()
steps = 0
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
for epoch_i in range(epoch_count):
for batch_images in get_batches(batch_size):
# Train Model
steps += 1
batch_z = np.random.uniform(-1, 1, size=[batch_size, z_dim])
batch_images = batch_images * 2
#Run optimizers
sess.run(d_train_opt, feed_dict={input_real: batch_images, input_z: batch_z})
sess.run(g_train_opt, feed_dict={input_real: batch_images, input_z: batch_z})
if(steps % 10 == 0):
#Get and print losses
train_loss_d = sess.run(d_loss, feed_dict={input_real: batch_images, input_z: batch_z})
train_loss_g = sess.run(g_loss, feed_dict={input_z: batch_z})
print("Epoch {}/{}...".format(epoch_i+1, epoch_count),
"Discriminator loss: {:.4f}".format(train_loss_d),
"Generator loss: {:.4f}".format(train_loss_g))
if(steps % 100 == 0):
show_generator_output(sess, 25, input_z, data_shape[3], data_image_mode)
MNIST¶
In [12]:
batch_size = 32
z_dim = 100
learning_rate = 0.0008
beta1 = 0.5
epochs = 2
mnist_dataset = helper.Dataset('mnist', glob(os.path.join(data_dir, 'mnist/*.jpg')))
with tf.Graph().as_default():
train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
mnist_dataset.shape, mnist_dataset.image_mode)
CelebA¶
In [13]:
batch_size = 32
z_dim = 100
learning_rate = 0.0008
beta1 = 0.5
epochs = 1
celeba_dataset = helper.Dataset('celeba', glob(os.path.join(data_dir, 'img_align_celeba/*.jpg')))
with tf.Graph().as_default():
train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
celeba_dataset.shape, celeba_dataset.image_mode)
